A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and Causality
Power forecasting models offer valuable insights on the electricity consumption patterns of clients, enabling the development of advanced strategies and applications aimed at energy saving, increased energy efficiency, and smart energy pricing. The data collection process for client consumption mode...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-09-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/14/19/6088 |
| _version_ | 1797516678092816384 |
|---|---|
| author | Dimitrios Kontogiannis Dimitrios Bargiotas Aspassia Daskalopulu Lefteri H. Tsoukalas |
| author_facet | Dimitrios Kontogiannis Dimitrios Bargiotas Aspassia Daskalopulu Lefteri H. Tsoukalas |
| author_sort | Dimitrios Kontogiannis |
| collection | DOAJ |
| description | Power forecasting models offer valuable insights on the electricity consumption patterns of clients, enabling the development of advanced strategies and applications aimed at energy saving, increased energy efficiency, and smart energy pricing. The data collection process for client consumption models is not always ideal and the resulting datasets often lead to compromises in the implementation of forecasting models, as well as suboptimal performance, due to several challenges. Therefore, combinations of elements that highlight relationships between clients need to be investigated in order to achieve more accurate consumption predictions. In this study, we exploited the combined effects of client similarity and causality, and developed a power consumption forecasting model that utilizes ensembles of long short-term memory (LSTM) networks. Our novel approach enables the derivation of different representations of the predicted consumption based on feature sets influenced by similarity and causality metrics. The resulting representations were used to train a meta-model, based on a multi-layer perceptron (MLP), in order to combine the results of the LSTM ensembles optimally. This combinatorial approach achieved better overall performance and yielded lower mean absolute percentage error when compared to the standalone LSTM ensembles that do not include similarity and causality. Additional experiments indicated that the combination of similarity and causality resulted in more performant models when compared to implementations utilizing only one element on the same model structure. |
| first_indexed | 2024-03-10T07:04:14Z |
| format | Article |
| id | doaj.art-d938c088fae34b0984e5a7f23580ea75 |
| institution | Directory Open Access Journal |
| issn | 1996-1073 |
| language | English |
| last_indexed | 2024-03-10T07:04:14Z |
| publishDate | 2021-09-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj.art-d938c088fae34b0984e5a7f23580ea752023-11-22T15:59:10ZengMDPI AGEnergies1996-10732021-09-011419608810.3390/en14196088A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and CausalityDimitrios Kontogiannis0Dimitrios Bargiotas1Aspassia Daskalopulu2Lefteri H. Tsoukalas3Department of Electrical and Computer Engineering, School of Engineering, University of Thessaly, 38221 Volos, GreeceDepartment of Electrical and Computer Engineering, School of Engineering, University of Thessaly, 38221 Volos, GreeceDepartment of Electrical and Computer Engineering, School of Engineering, University of Thessaly, 38221 Volos, GreeceAI Systems Lab, School of Nuclear Engineering, Purdue University, West Lafayette, IN 47907, USAPower forecasting models offer valuable insights on the electricity consumption patterns of clients, enabling the development of advanced strategies and applications aimed at energy saving, increased energy efficiency, and smart energy pricing. The data collection process for client consumption models is not always ideal and the resulting datasets often lead to compromises in the implementation of forecasting models, as well as suboptimal performance, due to several challenges. Therefore, combinations of elements that highlight relationships between clients need to be investigated in order to achieve more accurate consumption predictions. In this study, we exploited the combined effects of client similarity and causality, and developed a power consumption forecasting model that utilizes ensembles of long short-term memory (LSTM) networks. Our novel approach enables the derivation of different representations of the predicted consumption based on feature sets influenced by similarity and causality metrics. The resulting representations were used to train a meta-model, based on a multi-layer perceptron (MLP), in order to combine the results of the LSTM ensembles optimally. This combinatorial approach achieved better overall performance and yielded lower mean absolute percentage error when compared to the standalone LSTM ensembles that do not include similarity and causality. Additional experiments indicated that the combination of similarity and causality resulted in more performant models when compared to implementations utilizing only one element on the same model structure.https://www.mdpi.com/1996-1073/14/19/6088power forecastingenergymachine learningneural networksartificial intelligencedata analysis |
| spellingShingle | Dimitrios Kontogiannis Dimitrios Bargiotas Aspassia Daskalopulu Lefteri H. Tsoukalas A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and Causality Energies power forecasting energy machine learning neural networks artificial intelligence data analysis |
| title | A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and Causality |
| title_full | A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and Causality |
| title_fullStr | A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and Causality |
| title_full_unstemmed | A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and Causality |
| title_short | A Meta-Modeling Power Consumption Forecasting Approach Combining Client Similarity and Causality |
| title_sort | meta modeling power consumption forecasting approach combining client similarity and causality |
| topic | power forecasting energy machine learning neural networks artificial intelligence data analysis |
| url | https://www.mdpi.com/1996-1073/14/19/6088 |
| work_keys_str_mv | AT dimitrioskontogiannis ametamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality AT dimitriosbargiotas ametamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality AT aspassiadaskalopulu ametamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality AT lefterihtsoukalas ametamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality AT dimitrioskontogiannis metamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality AT dimitriosbargiotas metamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality AT aspassiadaskalopulu metamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality AT lefterihtsoukalas metamodelingpowerconsumptionforecastingapproachcombiningclientsimilarityandcausality |